SandboxAQ Pitches Quantitative AI Models to Energy Companies at ADIPEC

Hidary said these models, known as LQMs, complement large language models. LLMs work with text, while LQMs are built for complex scientific and engineering calculations. He said combining the two lets companies pair language processing with advanced scientific computation.

He described three uses for the energy sector. The first is AQCat, a model for designing new chemical catalysts. Hidary said such catalysts help turn raw materials taken from the ground into higher-value products further down the production chain. The second is AQ CFD, a computational fluid dynamics model that analyses how oil and gas move and flow, which he said can help companies understand and improve production and transport.

The third is carbon capture and storage. Hidary said SandboxAQ's work aims to turn captured carbon into products with higher economic value, rather than just storing it, and that this depends on developing advanced catalysts and industrial processes supported by its models.

Hidary praised how quickly the UAE is adopting AI. He pointed to the announcement that ADNOC will use OpenAI technology in its operations as a sign of how ambitious large energy companies have become about deploying AI models.

The interview reflects a theme of this year's ADIPEC: companies selling AI to the oil and gas industry beyond chatbots and office tools, aimed at core engineering and chemistry work. In the same week, ADNOC said it runs more than 200 AI applications and 70 robotics projects, and signed a multi-year agreement with OpenAI and G42.

Hidary did not tell WAM about any customer contracts in the UAE or the Gulf, and the report did not give pricing or deployment details for the models.